Esc2 promotes telomere stability in response to DNA replication stress

Abstract

Telomeric regions of the genome are inherently difficult-to-replicate due to their propensity to generate DNA secondary structures and form nucleoprotein complexes that can impede DNA replication fork progression. Precisely how cells respond to DNA replication stalling within a telomere remains poorly characterized, largely due to the methodological difficulties in analysing defined stalling events in molecular detail. Here, we utilized a site-specific DNA replication barrier mediated by the 'Tus/Ter' system to define the consequences of DNA replication perturbation within a single telomeric locus. Through molecular genetic analysis of this defined fork-stalling event, coupled with the use of a genome-wide genetic screen, we identified an important role for the SUMO-like domain protein, Esc2, in limiting genome rearrangements at a telomere. Moreover, we showed that these rearrangements are driven by the combined action of the Mph1 helicase and the homologous recombination machinery. Our findings demonstrate that chromosomal context influences cellular responses to a stalled replication fork and reveal protective factors that are required at telomeric loci to limit DNA replication stress-induced chromosomal instability.

Figure 1.

Site-specific replication fork stalling at Tus/Ter barriers causes localized mutagenesis. (A) Schematic diagram showing ChrVI and the position where the 21xTus/Ter barrier (incl. URA3 reporter gene) was inserted at either his2 or TEL06R (not drawn to scale). The red and green segments of Tus denote the restrictive and permissive faces, respectively. (B) Expression of Tus causes replication fork stalling at the TEL06R Tus/Ter barrier. Top panel depicts a diagram illustrating replication intermediates that can be visualized by 2DGE and Southern blotting. Genomic DNA was extracted at 75′ after G1 release, and BspHI-telomere fragments were analysed by 2DGE and Southern blotting using a URA3-specific probe. The white arrowhead indicates replication fork stalling at the Tus/Ter barrier in the cells expressing Tus. Cell cycle profiles are shown below. (C) URA3 and (D) CAN1 mutation rates were measured simultaneously for strains harbouring 21xTus/Ter at either his2 or TEL06R. Error bars indicate 95% confidence limits, and the numerical values above the columns indicate the fold increase in mutation rate between isogenic strains with (grey) or without (blue) Tus expression. Statistical analysis of differences in mutation rates was performed using a one-sided Mann–Whitney U test, and significance is indicated only when P < 0.01 (**P < 0.01; ****P < 0.0001). ns = difference not statistically significant.

Figure 2.

Esc2 suppresses URA3 mutagenesis and X-DNA formation at the TEL06R Tus/Ter barrier. (A) Summary of the genome-wide screen set-up (see Supplementary Figure S1C and ‘Materials and Methods’ for details). (B) The effect of ESC2 gene deletion on URA3 mutagenesis at the TEL06R (top) and his2 (bottom) Tus/Ter barriers. Data were analysed as described in Figure 1C.(C) WT, sgs1 and esc2 mutants harbouring the TEL06R Tus/Ter barrier were released from G1-arrest following induction of Tus. Genomic DNA was extracted at the indicated time points, and NheI-telomere fragments were analysed by 2DGE using a TEL06R-specific probe. The white arrowhead indicates replication fork stalling at the Tus/Ter barrier. Black arrows indicate unprocessed X-shaped DNA in sgs1 and esc2 mutants. Cell cycle profiles and quantification of the intensity of X-shaped DNA relative to that of the Tus-induced replication fork blocking spot are shown below. Quantifications were performed for three independent experiments, and error bars represent the standard deviation.

Figure 3.

The mutagenesis and X-DNA accumulation in esc2 mutants is dependent on HR. (A) The effect of deletion of RAD52 on URA3 mutagenesis at the TEL06R Tus/Ter barrier in an esc2 mutant. URA3 mutation rates were measured for the indicated strains harbouring the TEL06R Tus/Ter barrier. Data were analysed as described in Figure 1C. (B) Deletion of RAD52 suppresses accumulation of X-shaped DNA in esc2 mutants. Genomic DNA was extracted at 75′ after G1 release, and BspHI-telomere fragments were analysed by 2DGE and Southern blotting using URA3- and TEL06R-specific probes (two probes together). The black arrow indicates unprocessed X-DNA in the esc2 mutant. Cell cycle profiles and quantification of the intensity of X-shaped DNA relative to that of the Tus-induced replication fork blocking spot are shown below. Quantifications were performed for three independent experiments, and error bars represent the standard deviation. (C) The effect of deletion of SHU1 or EXO1 on URA3 mutagenesis at the TEL06R Tus/Ter barrier in an esc2 mutant. URA3 mutation rates were measured for the indicated strains harbouring the TEL06R Tus/Ter barrier. Data were analysed as described in Figure 1C.

Figure 4.

Replication fork stalling at the TEL06R Tus/Ter barrier triggers heterogeneous types of mutations. (A) Schematic illustrating TEL06R restriction fragments analysed by 1DGE (not drawn to scale). Dashed vertical lines indicate restriction enzyme sites. Probe 1 detects fragments 1 and 2. Probe 2 detects fragments 2 and 3. The positions of URA3 and the Tus/Ter barrier as well as the size of the restriction fragments are indicated. (B) DNA was extracted from individual 5-FOA resistant clones and AfeI-SalI-EcoNITEL06R fragments were analysed by 1DGE. Restriction fragments were visualized using either probe 1 (left panel) or probe 2 (right panel). WT clones are shown at the top and esc2 clones at the bottom. Markers (in kilobases in red text) are indicated on the leftmost portion of each gel. C represents non-mutated, WT restricted DNA (yielding ‘Fragment 2’ (11 kb) and ‘Fragment 1’ (6.7 kb + telomeric repeats) for probe 1. The corresponding fragments for probe 2 are ‘Fragment 2’ (11 kb) and ‘Fragment 3’ (4.7 kb). The asterisk denotes a non-specific band detected in all samples with probe 2. Note that with probe 1, the lower of the two bands (Fragment 1) in most esc2 clones is smaller than that in WT clones.

Figure 5.

esc2 mutants accumulate ‘truncations’ at the TEL06R Tus/Ter barrier. (A) Mutation types were identified at the TEL06R barrier in WT (left) and esc2 (right) strains. Pie charts indicate the relative proportions of mutation types, as indicated in the key below. Data were obtained by restriction digest analysis and telomere tailing/sequencing. (B) The specific mutation rate for individual types of mutations was calculated. n = number of times a specific mutation type was observed. Colour coding corresponds to pie chart in (A). (C) Schematic illustration of the mutation types (base errors, deletions within URA3 and URA3 truncations) identified by sequencing. (D) A specific type of chromosomal truncation is prevalent in esc2 mutants upon replication fork stalling at a telomere. Table showing the position (within the URA3 ORF) from where DNA sequence is lost and the number of clones displaying the given truncation. Data were obtained by telomere tailing and DNA sequencing. Black sequence is retained, whereas red sequence is lost. Note that 16 of 29 esc2 clones show an identical breakpoint (indicated by a black box).

Figure 6.

Abnormalities in esc2 strains are suppressed by deletion of MPH1. (A) Deletion of MPH1 suppresses URA3 mutagenesis at the TEL06R Tus/Ter barrier. Data were analysed as described in Figure 1C. (B) Deletion of MPH1 suppresses accumulation of X-shaped DNA in esc2 mutants. Genomic DNA was extracted at 75′ after G1 release, and BspHI-telomere fragments were analysed by 2DGE and Southern blotting using URA3- and TEL06R-specific probes (two probes together). The black arrow indicates unprocessed X-DNA in the esc2 mutant. Cell cycle profiles and quantification of the intensity of X-shaped DNA relative to that of the blocking spot are shown below. Quantifications were performed for three independent experiments, and error bars represent the standard deviation. (C) DNA was extracted from individual 5-FOA resistant clones and SalI-EcoNITEL06R fragments were analysed by 1DGE using probe 1 shown in Figure 4A. WT and esc2 clones are shown in the upper panels, while esc2 mph1 clones are shown in the lower panels. Markers (in kilobases) are as shown in Figure 4B. (D) Mutation types were identified at the TEL06R barrier in the mph1 (right) and esc2 mph1 (left) mutant. The pie charts indicate the relative proportions of mutation types, as indicated in the key below. Data were obtained by restriction digest analysis and telomere tailing/sequencing. (E) The specific mutation rate for individual types of mutations was calculated. n = number of times a specific mutation type was observed. Colour coding corresponds to pie chart in (D).

Figure 7.

Model for repair of a telomeric stalled replication fork. A replication fork is stalled at the Tus/Ter barrier (top). Resection of the lagging strand followed by DNA replication resumption would leave a ssDNA gap behind the fork (depicted on the left). The gap can be filled by post-replicative HR mediated by the Rad51 and Rad52 proteins amongst others. Deletions may arise via non-allelic HR repair of the ssDNA gap, in a process that is counteracted by Sgs1’s ability to disrupt aberrant strand invasion events. Alternatively, extended or prolonged fork regression might occur that is mediated by Mph1 (depicted on the right). Binding of Telomerase and the addition of telomeric repeats to the regressed fork could create a recombinogenic substrate that engages in HR with the telomeric repeats downstream of the Ter sites. Resolution of this HR intermediate would result in the observed telomere-proximal truncations observed in esc2 mutants. Large structural rearrangements could arise if unresolved HR intermediates are carried into mitosis and undergo chromosome breakage. Loss of Esc2 is proposed to lead to dysregulation of Mph1 activity, resulting in more extensive fork regression. Srs2 might normally serve to prevent HR intermediates from being processed correctly, with the result that srs2 cells funnel intermediates preferentially down the left branch of the pathway.